The hypothesis to be tested in that the high incidence of birth defects in infants of diabetic mothers is due to hyper-secretion of insulin by the fetal pancreas in response to maternal hyperglycemia in spite of insulin treatment of the mother. The incidence and type of birth defects will be determined on offspring of rats at very late stages of gestation or at birth and will be compared in the following groups: (1) offspring from normal mothers, (2) offspring from mothers with untreated diabetes which is produced by streptozotocin, (3) offspring from mothers whose diabetes is reversed by pancreatic transplantation prior to pregnancy, and (4) offspring from mothers with diabetes receiving insulin treatment. Techniques to be used include skeletal staining with alizarin red, whole embryo sectioning and new x-ray technique for skeletal details called the Faxitron. Prevention of birth defects through reversal of diabetes following pancreas transplantation or partial prevention from insulin treatment will provide a theoretical basis for a change in treatment of pregnant diabetics to prevent birth defects. Such an outcome may delineate the need for developing new techniques of treatment of diabetes such as transplantation or an artificial pancreas.